Open AccessNew structure-based models for the prediction of normal boiling point temperature of ternary azeotropes
Author(s) -
Zohreh Faramarzi,
Fatemeh Abbasitabar,
Hossein Jalali Jahromi,
Maziar Noei
Publication year - 2021
Publication title -
journal of the serbian chemical society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.227
H-Index - 45
eISSN - 1820-7421
pISSN - 0352-5139
DOI - 10.2298/jsc210218035f
Subject(s) - quantitative structure–activity relationship , ternary operation , boiling point , linear regression , test set , mathematics , predictability , training set , thermodynamics , computer science , chemistry , statistics , artificial intelligence , machine learning , organic chemistry , physics , programming language
Recently, development of the QSPR models for mixtures has received much attention. The QSPR modelling of mixtures requires the use of the appropriate mixture descriptors. In this study, 12 mathematical equations were considered to compute mixture descriptors from the individual components for the prediction of normal boiling points of 78 ternary azeotropic mixtures. Multiple linear regression (MLR) was employed to build all QSPR models. Memorized_ ACO algorithm was employed for subset variable selection. An ensemble model was also constructed using averaging strategy to improve the predictability of the final QSAR model. The models have been validated by a test set comprised of 24 ternary azeotropes and by different statistical tests. The resulted ensemble QSPR model had R2 training, R2 test and q2 of 0.97, 0.95, and 0.96, respectively. The mean absolute error (MAE), as a good indicator of model performance, were found to be 3.06 and 3.52 for training and testing sets, respectively.